Research And Application On Generalized Predictive Control Based On Particle Swarm Optimization Algorithm

The dissertation focuses on generalized predictive control based on particle swarm optimization algorithm. On the basis of the study of generalized predictive control technology, particle swarm optimization algorithm and propose an improvement aimed have analyzed at the deficiency of the iterative optimization. Integrating particle swarm optimization algorithm with gradient optimization method, a hybrid algorithm based on multimodal has been proposed and applied to control component contents in the process for extraction of the rare-earth extraction. The following aspects are studied.Firstly, basic generalized predictive control algorithm and it’s improved implicit algorithm have been studied. Because implicit algorithm avoid large amounts of computation which from solve Diophatine equation, so reduce the computation time. The simulation shows that the algorithm have good control performance. Moreover, the adjustable parameters have been discussed. Simulation result shows the effect of the parameters on the system performance, and the general principle of parameters selection is obtained.Secondly, generalized predictive control based on particle swarm optimization has been designed. As for the traditional generalized predictive control technology, the iterative optimization turned to gradient optimization method, with the deficiency of large time consuming, complicated process, large computation, as well as local optima. Particle swarm optimization is an algorithm which originates from swarm intelligence theory. The algorithm contains advantages including fast convergence, high accuracy and easy implementation. Integrate particle swarm optimization algorithm with gradient optimization method, and design a hybrid algorithm based on multimodal. Consequently, it obtains the optimal control input.Finally, the algorithm proposed in this dissertation is applied to control component contents in the process for separation of the rare-earth extraction. Based on the modeling of the process of component contents control, the system regulates the flow rate of extracted liquid material and liquid twist and realizes the tracking of the component contents. Simulation shows the tracking performance of the method.In the current dissertation, the generalized predictive control based on particle swarm optimization algorithm have been used in the system of rare-earth extraction. Theory study and simulations analyses are presented to show the performance of the method proposed. This work has certain value for the production of rare-earth.